Within for every website, i obtained step 1,100 pages regarding heterosexual boys and 1,one hundred thousand users from heterosexual lady. Users search for profiles via geographical place, ages, and you may gender filter systems.

To ensure a geographical dispersion out-of pages, we picked equal amounts of users out-of five big metropolises as well as urban, suburban, and you will rural section: La, Denver, Chicago, Atlanta, and you can New york. We randomly picked zero rules off each of the five parts to look for users. Inside for each and every area code, for each gender, we up coming randomly picked users certainly one of four age groups: very early more youthful adulthood (aged 18–31 ages; n = step 1,000; M = , SD = step 3.17), later more youthful adulthood/early midlife (old 29–44 age; n = step one,000; Meters = , SD = 5.61), late midlife (aged fifty–64 age; n = step one,000; Yards = , SD = step three.99), and you will the elderly (aged over 65 ages; letter = step one,000; Yards = , SD = 4.29). I utilized this type of stratifications in order to guarantee the full age range off matchmaking pages when you look at the testing. Because the elderly group you will incorporate up to three decades, i addressed years just like the a continuing adjustable unlike just like the an effective grouping changeable inside analyses.

Of each character, we extracted: gender, ages, ethnicity, and also the “Throughout the Me” otherwise “In my Conditions” totally free impulse section. To assure privacy out-of character writers, i failed to receive extra market pointers (age.grams., studies, religious tastes, income) which could act as determining pointers. The testing method is portrayed in the Additional Appendix Good .

The try ranged during the years out-of 18 to 95 many years. An independent t-try shown zero difference in suggest decades for ladies (Meters = , SD = ) and you may boys (M = , SD = ). The new summary of ethnicity on the sample are 70% White/Caucasian, 11% Black/Dark colored, 7% Hispanic/Latino, 2% Far-eastern, and you will 10% blended competition/most other.

Character content

We used the LIWC application to analyze the content of your users. This program exercise the fresh new regularity and dimensions of particular categories of conditions within this a book file. The fresh LIWC program measures up each word-of a book file having an inside dictionary of greater than cuatro,500 terminology assigned to keyword classes. This study received for the 11 oriented LIWC categories: first-individual just one pronouns, first-individual plural pronouns, family relations, nearest and dearest, performs, conclusion, currency, fitness, sexuality, self-confident emotion, and you will negative feelings. Table 1 includes example terms for the all the LIWC group (for additional information regarding these types of rules, look for LIWC, 2007).

We plus written a category of words to have attractiveness not available inside built LIWC kinds. We used strategies getting construction out-of LIWC classes ( Tausczik & Pennebaker, 2010) of the promoting an extensive set of terms out-of dictionaries, thesauruses, forms off past browse, and you may terminology from the study group. Up coming, i selected twenty-five terminology most associate regarding appeal centered on physical appearance inside thesauruses and you can participant solutions (age.grams., lovable, good-looking, good looking, hot). This new elegance classification are nearly entirely distinct from the fresh new sexual classification, in just you to definitely overlapping term (sexy). Instances on appeal classification are also utilized in Dining table step one; to the complete range of terminology regarding the elegance category, look for Additional Table 1 .

Analytical Approach

We basic tested detailed statistics towards dimensions of terminology anybody various years included in the matchmaking pages. I along with made illustrative data depicting the most used conditions.

I after that looked to theory assessment having fun with typical minimum squares regression. The outcomes details within this study was indeed the new proportion out-of terms and conditions suitable all the several classes on LIWC analyses. The latest LIWC classes was in fact every certainly skewed due to the matter out of no opinions (we.age., participant failed to explore any terms regarding the category). I ran for every analysis which have a rectangular-means transformation (familiar with address low-normality for the past education utilizing the LIWC; DeAndrea, Shaw & Levine, 2010; Hirsh & Peterson, 2009). The fresh trend out of results try comparable shortly after applying the changes. To own simple interpretation, results is actually shown with the untransformed LIWC classification study. This new independent changeable is actually years, handled once the a continuing adjustable. I and additionally provided sex.